--- Log opened Wed Jan 09 00:00:15 2019 02:30 -!- TamasBlummer1 [~Thunderbi@p200300DD67243B0810A52DBE0F92F0F8.dip0.t-ipconnect.de] has joined #rust-bitcoin 02:32 -!- TamasBlummer [~Thunderbi@p200300DD67243B7010A52DBE0F92F0F8.dip0.t-ipconnect.de] has quit [Ping timeout: 264 seconds] 02:32 -!- TamasBlummer1 is now known as TamasBlummer 03:07 -!- grubles__ [~grubles@unaffiliated/grubles] has quit [Remote host closed the connection] 07:16 -!- michaelsdunn1 [~michaelsd@unaffiliated/michaelsdunn1] has joined #rust-bitcoin 07:29 < BlueMatt> real_or_random/andytoshi: hmmmm do we normally merge without merge commits? 07:39 -!- icota_ [~igor@141-136-181-157.dsl.iskon.hr] has quit [Quit: Konversation terminated!] 07:39 -!- icota_ [~igor@141-136-181-157.dsl.iskon.hr] has joined #rust-bitcoin 08:17 -!- grubles__ [~grubles@unaffiliated/grubles] has joined #rust-bitcoin 08:34 < real_or_random> BlueMatt: oh, I used that github squash feature for the first time, I thought it just squashes all PR commits into one and then merges normally. but it even leaves out the merge commit 08:34 < BlueMatt> yea 08:34 < BlueMatt> oh well 08:34 < BlueMatt> nbd 08:34 < BlueMatt> it still includes the pr # in the commit msg, so I'm not too worried 08:34 < BlueMatt> and it gets rebased, apparently 08:34 < BlueMatt> just noting that it seems like we dont usually do that 08:35 < real_or_random> indeed 08:38 < BlueMatt> real_or_random: as for fuzztarget, the only obvious thing I can come up with is a #[doc_hidden]pub static REALLY_ENABLE_BROKEN_CRYPTO: bool = false; that fuzzers have to set 08:38 < BlueMatt> and then just assert it in the fuzztarget functions 08:38 < real_or_random> have you seen my latest comment? 08:39 < BlueMatt> oh, no, i hadnt 08:39 < BlueMatt> hmm, i guess we could do that, too, you mean in the build script? 08:40 < BlueMatt> I'm game with either, just as long as we can include it in the crates version and it gives a reasonable "you forgot to do X" result when you dont do it 08:40 < BlueMatt> env var in build.rs may actually generate a better error message 08:46 < real_or_random> I think it will be best to require the env variable in build.rs AND whenever a fuzzer function is called 08:46 < BlueMatt> could do that too, though that may be overkill 08:47 < real_or_random> so you're saying the build is enough 08:48 < BlueMatt> and checking env variables lots may slow down the fuzzer 08:48 < BlueMatt> I'd think so? 08:48 < real_or_random> hm okay, yes. we could read it once startup though 08:49 < BlueMatt> is there a way to call a function at startup in rust? 08:49 < stevenroose> BlueMatt, andytoshi: is there any code for compact blocks in rust-bitcoin yet? 08:49 < BlueMatt> in a library, that is 08:49 < BlueMatt> my recollection is no 08:49 < BlueMatt> stevenroose: not afaik? why do you want it? 08:51 < BlueMatt> stevenroose: note that downloading blocks over compact blocks doesnt get you the protection of your full node pre-validating 08:51 < BlueMatt> cause it will relay spv-validated things 08:52 < real_or_random> yeah okay, I only know https://docs.rs/lazy_static/1.2.0/lazy_static/ that adding a dependency is overkill 08:52 < real_or_random> *but 08:53 < BlueMatt> yea, plus an atomic check at each entry point 08:53 < BlueMatt> which, sure, isnt that expensive on x86, but on other platforms more so 08:58 < real_or_random> agreed 09:00 < real_or_random> for reference, I've also found this now https://docs.rs/ctor/0.1.6/ctor/ ... but yep, we should stick to the build-time check 09:02 < stevenroose> BlueMatt: we just introduced it in Elements/Liquid for relaying between signers. And currently we added 2 or 3 rpc calls to elements that are stateless 09:02 < BlueMatt> real_or_random: I presume it does it via a c-lib which does the __attribute__(constructor)) stuff, or something like that 09:02 < real_or_random> yep 09:02 < stevenroose> and since the signer code uses rust already, we shouldn't need those rpcs if the compact blocks calls can be done in rust locally 09:29 < real_or_random> BlueMatt: argh env variables are kind of apparently https://github.com/rust-lang/cargo/issues/4587 09:47 < BlueMatt> real_or_random: huh? is that saying it wont rebuild if you change the env? 09:48 < BlueMatt> real_or_random: I assume we can just report an error that tells you to cargo clean first 10:10 < real_or_random> well if build.rs is not run, then we can't report an error. the bad scenario is that you configure fuzztarget, set the env variable and build. and then you reboot, and build again... env not set but cargo build will return normally. this kinda defeats the purpose 10:10 < BlueMatt> ah 10:11 < BlueMatt> well, I dunno, how worried are we about devs doing this? 10:11 < BlueMatt> I thought we were more worried about people who just blindly turn on features without knowing what they mean? 10:12 < BlueMatt> its not like we'd use the env/build.rs to actaully compile the fuzz branches 10:13 < real_or_random> yeah maybe you're right and this is just good enough 10:13 < BlueMatt> I mean it definitely *helps* 10:13 < real_or_random> yep 10:13 < BlueMatt> even if it isnt complete 10:45 < BlueMatt> ariard: re: your comments here from yesterday: yes, I suppose you're right 10:45 < BlueMatt> ariard: as long as test we can re-claim-from-revoked-htlc-tx-output we can fail-backwards immediately 16:02 -!- michaelsdunn1 [~michaelsd@unaffiliated/michaelsdunn1] has quit [Remote host closed the connection] 20:21 -!- Guest72 [~Guest72@CPE90724009c0f9-CMac202ec308b0.cpe.net.cable.rogers.com] has joined #rust-bitcoin 20:35 -!- Guest72 [~Guest72@CPE90724009c0f9-CMac202ec308b0.cpe.net.cable.rogers.com] has quit [Quit: Guest72] 20:38 -!- midnightmagic [~midnightm@unaffiliated/midnightmagic] has quit [Ping timeout: 260 seconds] 20:39 -!- grubles__ is now known as grubles 20:43 < ariard> BlueMatt: still not done with #284, writing test to find some dumb dust cases 20:49 -!- midnightmagic [~midnightm@unaffiliated/midnightmagic] has joined #rust-bitcoin 22:19 -!- midnightmagic [~midnightm@unaffiliated/midnightmagic] has quit [Ping timeout: 264 seconds] 22:33 -!- midnightmagic [~midnightm@unaffiliated/midnightmagic] has joined #rust-bitcoin --- Log closed Thu Jan 10 00:00:16 2019